Antique Furniture

and the impact of its environment

Mike Podmaniczky


Dressing table incorporating 'breadboard' construction  
Close-up of dressing table showing orientation of leg post to side panel  
Dressing table with leg post oriented at 90 degrees to the grain of the side. ‘Cross-grain’ or ‘breadboard’ construction often results in damage to the wide board caused by environmental fluctuations.  

There is a common, intuitive understanding that wood furnishings physically respond to changes in the surrounding environment. When the relative humidity rises, close-fitting cabinet doors or drawers may stick. When the humidity drops, the problems clear up. If it drops a great deal, old cracks in door panels open up and occasionally new cracks appear. We have all developed an appreciation that wood moves with the weather. Under certain circumstances, this movement can manifest itself as more significant damage in fine old furniture or, indeed, in brand new furniture constructed with traditional materials and techniques but without due regard for timber’s responsiveness to environmental changes.

In an enclosed, and nominally climate-controlled space, the question of environmental impact on furniture, and indeed any wood objects, hinges on two characteristics of timber: good news and bad news, if you will. The good news is that wood is an elastic material that can absorb a great deal of stress without undergoing permanent deformation. The bad news is that under certain, quite commonly occurring conditions, wood will respond to stress as a plastic material, meaning that it will retain the new shape it has been forced into and be permanently deformed. It should also be noted that there is a point in the application of stress beyond which strain can no longer be accommodated with elastic or plastic deformation, and the wood fails. In these cases, wood breaks, splits, or under extreme conditions, crushes.


An archer bends his yew bow and, time and time again, it snaps back to its original configuration. A small boat builder takes long strips of oak or ash, and permanently bends them into the shapes he needs for framing small watercraft. The latter may or may not require heat (commonly provided by steam) depending on the sharpness of the bend and the moisture content of the wood. However, as any archer knows, if the bow is left strung, and thus bent, for any length of time, it will take on a permanent bend that will not return to its initial shape even if the stress of the bowstring is relieved. Conversely, a wood element that has been forced into a new shape, perhaps with the application of heat, will not hold that shape unless it is held to it for a period of time. Rather, it will be seen to slowly relax back toward its initial, pre-stressed shape.

The obvious variable in these examples is time. Time is a key element in the measuring of viscosity, and wood’s elastic or plastic reaction to stress makes it a visco-elastic material. Under long-term, externally applied forces, wood will ‘flow’ to the conformation it is pushed to. This is the result of the sub-cellular components of wood sliding in relation to one another and is made possible by the water present in the wood’s cellulosic structure. If the molecular water present in wood that is in equilibrium with the surrounding environment is allowed to fluctuate, with water coming and going within the cellulosic structure, the slippage is enhanced and plastic deformation hastened and exacerbated. Some furniture-making processes actually take advantage of this phenomenon. For example, bentwood chair elements can be made from green, unseasoned wood that is forced to shape and allowed to dry under restraint. The massive loss of both liquid and molecular water accompanying the drying process is accompanied by substantial internal, sub-cellular rearrangement and the result is a wood element that has relaxed to a deformed shape. Again, time is a critical factor.

It is a rare occurrence that a piece of furniture is subjected to this kind of treatment, although a book shelf, for example, heavily loaded and supported only at the ends, will creep or flow to the bowed shape on account of the weight of the books. Induced deformation such as this will take a given amount of time under static environmental conditions. However, when the loaded shelf is subjected to varying levels of moisture content, such as when the relative humidity in the surrounding environment fluctuates, the deformation greatly increases. This is known as the mechano-sorptive effect and is the result of the flexing and separation of layers of cellulosic structure on a submicroscopic level. As molecular water forces its way into the structure at high humidity, and then works its way out, the wood substance is able to more readily rearrange itself and creep to a new conformation. This results in a more rapid and exaggerated deformation each time there is a humidity change than would be the case under relatively static environmental conditions.

Of course, variable environmental conditions are the norm. Fortunately, however, except for the occasional overloaded bookshelf, most furniture is not under the same kind of applied stress as the archer’s bow while the humidity cycles and moisture content of the wood changes.


There is a much more pernicious manifestation of stress or strain caused by environmental conditions that is related to ‘inherent vice’. This is a term used by conservators for problems with an object that are inherent in the materials it is made from or how those materials are used. In the case of furniture, there is an inherent vice due to the nature of wood itself and this is exacerbated when it is put into an orientation that pits one element against others in the overall assembly. In this case, individual elements may be assembled into an unstressed piece of furniture, but when it is subjected to varying humidity, the dimensional changes in one element can be resisted by others in the surrounding structure.

  Damaged marquetry and veneer decoration on flush, panel-in-frame door
  Interior face of flush, panel-in-frame door
  Marquetry and veneer decoration applied over flush, panel-in-frame door: the uncoloured repairs trace the seam between the frame and the panels where differential movement has resulted in a history of damage ‘telegraphed’ through the decorative veneers (seen in close-up below right)

Wood is an anisotropic material, which means that it demonstrates different physical qualities depending on the orientation of observation or use. (Clay, in contrast, is an isotropic material.) It is also hygroscopic in that there is a chemical attraction between cellulose and water. Timber reacts to moisture variation by taking up or giving off molecular water and expanding or contracting across the grain, but not along the grain. A piece of timber may become wider or narrower in response to moisture content changes, but it does not significantly change in length. If two pieces of timber are assembled to each other with their grain direction oriented at right angles and then exposed to humidity variations, there will be unequal dimensional change at the interface between the two. This cross-grain construction is the fundamental inherent vice associated with furniture.

Furniture makers have always had to deal with this phenomenon and construction practices have developed to accommodate it and to minimise its negative effects. Unfortunately, in many cases, craftsmen have chosen to be cavalier with their approach to building, or there simply is no way to avoid the problems caused by the anisotropy of the material.

The panel-in-frame is a classic example of accommodating anisotropy and humidity-induced dimensional changes. Here, a narrow frame is constructed with the grain of its elements running in line with the perimeter. Therefore there will be no change of the overall frame dimension with the changing weather. The opening is filled with a thin panel of wood that is set into a groove running around the inside perimeter of the frame. The panel is never fitted tightly or fastened in any way, so that it is free to expand or contract without having an effect on the external dimensions of the assembly. In this way, a practical, stable panel is constructed that can readily be incorporated into a piece of furniture or wall panelling.

The sharp eye will note that, at the corners of the frame, the grain directions of the two elements are at 90 degrees to each other. However, here the scale of the cross-grain construction is small enough to accommodate differential movement. These elements are usually mortise and tenon joints, fixed with a pin. At this dimension, wood movement is barely detectable, and the tenon is relatively free to move within the mortise. One might think of this as being within the elastic range of wood movement. However, while the effects of cross-grain construction in a chair rail and leg joint are insignificant, they become increasingly magnified as the dimension of the joint increases, for example at the interface of a wide dressing table side with a long leg post (see illustrations at top of page). Here, the chance of problems dramatically rises.

Breadboard construction, usually found in table tops, is perhaps the best example of the inherent vice that can be established with poorly designed building techniques. The damage that can result from this is easily recognised. This type of construction uses narrow timber battens set cross-wise at either end of a wide board or multiple edge-joined boards. The battens are intended to hold the wide boards flat and are presumably used in the vain hope that the overall panel might be kept from changing shape. When such a wide board either shrinks or swells, the end battens do not change length and the difference between board and batten is immediately apparent. Hypothetically, if the joint between the board and the batten is not restrictive, as humidity rises or falls the board will either stand out proud from the ends of the batten, or vice-versa. (Just as a broken clock is right twice a day, the alignment of board and batten will be correct at the relative humidity that maintained when the panel was originally assembled.)

But this is more than an issue of dimensional inconsistencies. If the assembly is not properly executed, the result can be a split table top. If the batten is tightly affixed to the end of the table to prevent differential movement and the top board tries to shrink in response to lowered relative humidity, a great deal of stress, in this case tension, is applied across the grain. As timber is much stronger in compression across the grain than in tension (think splitting firewood versus putting a car up on wooden blocks) there will be little margin for this to cause plastic deformation of the top, and since it won’t stretch across the grain, it will split.


The next consequence of inherent vice is perhaps the most interesting, and also the least commonly understood. The healthcare profession refers to hypertension as ‘the silent killer’ since there are no symptoms until the damage is done; compression set shrinkage in wood is comparable.

Consider once more the case of breadboard construction in a high-humidity environment. The wide board will try to swell, but if the end batten is tightly fixed to it movement will be restrained, effectively exerting the equivalent of compression stress across the grain. Restrained, there is no apparent movement or change in the width of the board, but ‘expansion’ still takes place within the wood, on a microscopic level. The wood cells deform only slightly, but the cumulative effect is significant and irreversible. This is referred to as compression set shrinkage. It would take tremendous force applied externally to crush wood but, under the conditions described, it happens routinely. When the humidity declines, the board will attempt to shrink, but because of permanent damage to the cell structure as a result of compression set shrinkage, the overall dimension of the board will be less than it was originally. Because it is still restrained by the cross-grain end battens, the result will be that the board splits, generally near the centre of the board.

Damaged marquetry and veneer decoration

It is quite common to see splits in the sides of old pieces of furniture, and it is generally ascribed to ‘shrinkage’ over the however-many hundred years since the object was made. It is likely that most furniture suffered the initial, internal damage of compression set shrinkage early on in life but was not forced to a low enough moisture content to cause the splitting until sometime in the 19th century with the advent of central heating and the very dry conditions associated with it.


Decorative veneer and marquetry adds further opportunity for environmentally caused damage in antique furniture. Thin layers of wood, either in solid sheets of veneer, or jigsaw puzzle-like images assembled from cut out pieces of veneer are traditionally applied to a solid timber substrate with hot animal glue, a water-based adhesive. Direct contact from liquid water risks the detachment of the veneers, but they are more likely to be disturbed by fluctuations of relative humidity, particularly if the grain direction of the veneer is in opposition to the grain of the solid wood substrate. When the substrate is made up of a number of solid wood elements, themselves set in directional opposition to each another, the resulting conflict of movement can ruin the applied decoration. Although not strictly an issue of compression set shrinkage, any conflicting movement of a substrate will be ‘telegraphed’ through the veneer on the surface, sometimes even tearing the delicate applied decoration. The inherent vice of this type of furniture decoration makes it particularly susceptible to environmental impact.


  Historic Georgian interior
  Window blinds filter the light to protect Georgian furniture at No 1 Royal Crescent, Bath (Photo: Jonathan Taylor, by kind permission of The Bath Preservation Trust)

The somewhat forgiving nature of wood means that, within reason, furniture can be maintained in a relatively healthy state unless subjected to environmental extremes. The ideal environment (rarely maintained) for furniture is 18°C (65°F), and 50%RH. As desirable as the ideal is, environmental fluctuations need not result in damage to furniture. As stated, before the age of central heating, most furniture survived seasonal changes relatively well, primarily because they happened over time. Buffering an interior environment from radical swings in temperature and humidity allows the visco-elastic nature of wood to accommodate to the changes over a period of time and greatly reduces the risk of environmentally caused damage to furniture.

This article is reproduced from The Building Conservation Directory, 2008


MIKE PODMANICZKY is a furniture conservator who spent most of his career at Winterthur Museum in Delaware in the
US. He now teaches furniture making and conservation at West Dean College. Prior to earning his MA degree in conservation through the Smithsonian Institution in Washington, DC, he worked as a pattern maker, boat builder and furniture maker.

Further information



Joinery and Cabinetmaking



Antique and furniture restoration

Cabinet makers

Fine joinery

French polishers

Wood carvers and turners
Site Map